October 21 2009

The New-Wave of Internet Search

The real-time environment of the internet has evolved a concerning dichotomy for fact and fiction. When Michael Jackson died in June, word spread too fast for Google to cope, and the site began blocking any search for “Michael Jackson”. But the lust for up-to-the-minute news could not be kept at bay - word spread through tweets and other micro-blogs regardless. Real time reporting had the edge.

Or did it? On the very same day, Jeff Goldblum was also reported dead, provoking a similar unstoppable surge of rumours and gossip

Jeff Goldblum and Micahel Jackson

Real-time internet may be powerful in keeping up-to-date with news, but it most certainly lacks reliability – Jeff Goldblum, as it turns out, is not dead after all. Therein lies the problem. In the explosion of information surrounding extremely recent events, how can we distinguish fact from fiction when we don’t know how the fuse was lit?

Twittertown

Real time is the talk of the internet search town at the moment. Twitter, the biggest contributor to real-time data, continues to grow in popularity and Twitter Search, the only real-time search engine with access to all tweets, is in a powerful position. But a wave of search engines which pull together data from across the web have sprung up recently. Sites like Collecta, OneRiot, and Scoopler broaden real-time search to include blogs, articles, photos, and videos as well as tweets. And in recent months the big players have shown that they want a piece of the action too: Google, Bing and Facebook have all taken steps to keep up with the real-time crowd.

But what exactly is real-time search and why is everyone so excited about it? Traditionally, search engines like Google have organised their results based on authority. Sites have authority if they have grown slowly and organically over time. Real-time search engines, on the other hand, sort their results by how recent they are. Through these search engines, users can access a river of the latest information on whatever topic they choose.

Collecta

Increasingly people are turning to the web to find out what is happening right now – the recent protests in Iran are a perfect example (and a frequently mentioned one). But when you search for a term in a traditional search engine the results look very similar day after day. If a volcano is erupting, followers on the web do not want to read an old article about the properties of lava, however authoritative it may be. With the real-time web your results will be different every time, and often refresh before your eyes. So no out-of-date articles, and no need to wait for news; users have access to up-to-the-minute comments and images. They can find out what is happening at the heart of a demonstration or at the site of a volcanic eruption as the event is taking place.

The real-time web also tells you what topics everyone’s talking about.  Most real-time search engines display trending topics, the most popular at that moment, and many can sort results by categories such as sport or entertainment. Want to know what your colleagues will be talking about at work tomorrow? A real-time search will probably tell you.

But a real-time search will probably also tell you all the information you didn’t want to know, or didn’t care about. Aside from rumours becoming gospel faster than you think possible, the current main disadvantage of real-time search engines is their inability to filter unwanted messages or irrelevant noise from results. The river just keeps on flowing regardless of what it has picked up along the way.

Collecta have openly stated they are currently paying no regard to relevancy in their results. Oneriot, however, have begun to experiment with reliability by introducing Pulserank, a toolbar which not only takes into account the freshness of the information, but also the authority of the website and person posting the information, alongside the velocity of the information passing around the whole web. The potential for the tool is huge, but although this seems like a reasonable approach, it may not catch something important as fast as simply watching the unadulterated stream.

Oneriot

Although far from fully effective, the Pulserank toolbar does pave the way for the necessary filters which real-time searching will require as the phenomenon grows. More users will undoubtedly lead to more spam and more noise being generated, increasing the need for an effective filter barrier. The challenge for real-time search engines is to combine recency, relevancy and reliability in their results without becoming elitist and losing the organic chatter of the online crowd.

Problems aside, the current animation surrounding the technology should lead to exciting developments. One such possibility being the use of real-time internet searching as an alert system – by signaling variations in the stream of mentions for a particular query, any abnormal rise in the quantity of chatter would trigger a notification. So the future of real-time search is bright, if hazy. Entrepreneur Edo Segal believes that old-school search will never vanish, but real-time news will create a society where we have an omnipresent sense of the moment.

Post your comment

Comments

  • Its very important

    Posted by hotspot shield, 01/07/2013 9:50am (4 years ago)

  • Valuable thoughts and advices. I read your topic with great interest.

    Posted by Arsento, 01/07/2013 9:50am (4 years ago)

  • I read a few topics. I respect your work and added blog to favorites.

    Posted by Floost, 01/07/2013 9:50am (4 years ago)

RSS feed for comments on this page | RSS feed for all comments